Big Data for Big Problems

watson_computer

 

In Isaac Asimov’s short story “The Last Question,” the year is 2061, and for decades a computer called Multivac has been helping humanity design the ships that enable man to reach the Moon, Mars, and Venus. By May of that year, with Multivac’s help, all of Earth is run by “invisible beams of sunlight”. This incredible super-computer is described as being miles wide, and self-adjusting and self-correcting. But the scientists who maintain it realize that despite the seemingly infinite amount of energy that the sun can provide, it will not last forever. Someday the sun will burn out. Someday all of the stars will burn out. Entropy will destroy man’s universe.

So, one drunken night, the scientists ask the computer, “How can the net amount of entropy of the universe be massively decreased?”

Multivac replies: INSUFFICIENT DATA FOR MEANINGFUL ANSWER. The drunk scientists run off, and forget about the question. Yet trillions of years later, and multiple iterations of Multivac into the future, the question is still being processed. Multivac has extended into hyperspace, a network of post-human consciousnesses. When Man’s last mind finally fuses with the Multivac, it finds the answer:

“The consciousness of AC encompassed all of what had once been a Universe and brooded over what was now Chaos. Step by step, it must be done. And AC said, ‘LET THERE BE LIGHT!’ And there was light —” [1]

Asimov wrote this story in 1956, as an extrapolation of the vision in the 1950s of a centralized global computer. Today, it seems we might be well on the way to Asimov’s vision, though it certainly won’t be the miles-wide machine he foresaw. In 2014 IBM announced that one of its supercomputers, Watson, would be delivering consulting services to address development obstacles in Africa.[2] While this is not quite as fantastic as designing a spaceship to Mars and Venus, it does signify an astounding leap in technology.

In only six years, Watson progressed from a mere idea to the world champion of the quiz show game Jeopardy!.[3]After winning Jeopardy! IBM decided to implement the natural language processing, hypothesis generation, and evidence-based learning powers of the computer to solve problems ranging from healthcare, education, transportation, and agriculture. Using mobile devices to access the cloud-based Watson, people in Africa will be able to utilize the supercomputer’s massive data processing capabilities by asking it intuitive questions in everyday speech.

This kind of “smart” data mining has already borne fruit in Morocco, literally: using huge amounts of data, Moroccan farmers were able to improve how crops are grown by predicting weather, demand, and disease outbreaks. A delivery firm in Lagos, Nigeria is currently using the Watson system to improve delivery times and schedules. Analytics could provide the state information about the quality of transportation in highly congested cities. Healthcare professionals could use the system to diagnose illnesses and prescribe proper medication, as some are doing already.[4] The BBC also suggests that Watson could “help discover why sub-Saharan Africa currently accounts for 22% of all cervical cancers. It could suggest new ways to treat and prevent the disease.”[5]

This project is named Lucy, after the earliest known human ancestor fossil, and will cost 100 million dollars with around ten years to complete. IBM scientist Uyi Stewart, imagines  that Watson will transform Africa in the way that mobile banking has already. Despite this huge potential, however, IBM has only made 100 million dollars form Watson in the past three years. Nevertheless, IBM holds fast to the common belief that Big Data is the next big thing in business, and has invested one billion dollars into commodifying Watson’s services.

Besides downplaying the risks involved in speculating about the next big thing in business, mainstream media have failed to address other core aspects of the Watson phenomenon. From a technical standpoint, the articles do not mention the actual user experience of the Watson system. How do people in Africa connect to the service reliably? How long does it take to process and deliver a response? What languages will be supported? How much will it cost to access the service? The articles so far fail to mention these technical aspects, and their rhetoric is not skeptical enough of the full capabilities of the machine. It is as though, having won Jeopardy!, the machine is now infallible.

Even more worrisome are the sociological impacts of the technology that the articles seem to have ignored. For instance, there does not seem to be a limit to what you can ask Watson. It is conceivable that people may use the system maliciously, perhaps using Watson to discover vulnerabilities in a city’s power grid or water system. It is also possible that, since the vast amount of Wikipedia, from which Watson gleans its knowledge, has a Western bias, Watson may provide some answers that are culturally incompatible for some of its users.

But perhaps most surprising is that the articles don’t mention that Watson is almost outdated already. It certainly does not even make the list of the top 500 supercomputers, and in the time since it has won Jeopardy! in 2011, it has shrunk 90 percent—“from the size of a master bedroom to three stacked pizza boxes,” and has experienced a 2,400 percent improvement in performance, which indicates that larger, smarter computers are certainly available.[6] The fact that this technology has experienced such amazing progress in such short time must be worthy of some kind of mention for the reader to get an accurate grasp of the rate of change and realm of possibilities involved in the world of computing.

Despite these shortcomings, the mainstream media has at least done a fair job of avoiding alarmist headlines. It is easy for technophobes to have the last say, but this doesn’t seem to be the case with Watson. In fact, the main issue is that mainstream media might be overly optimistic about the ability of technology to solve all of humanity’s problems.

But maybe this optimism is not just unique to the media. Maybe we all want to believe, like Isaac Asimov did, that if we keep trying, we will someday build a computer that can tell us how to massively reverse entropy, and save humanity from the ultimate futility of the cold dead chaotic universe. Maybe we all just hope that computers will be the ones to save us from The End.



[1] http://www.physics.princeton.edu/ph115/LQ.pdf

[2] http://www.reuters.com/article/2014/02/06/us-ibm-africa-idUSBREA1507H20140206

[3] http://www.nytimes.com/2011/02/17/science/17jeopardy-watson.html?pagewanted=all&_r=0

[4] http://www.forbes.com/sites/bruceupbin/2013/02/08/ibms-watson-gets-its-first-piece-of-business-in-healthcare/

[5] http://www.bbc.co.uk/news/technology-26065991

[6] http://www-03.ibm.com/press/us/en/pressrelease/42869.wss

2 Responses to “Big Data for Big Problems”

Leave a Reply

Basic HTML is allowed. Your email address will not be published.

Subscribe to this comment feed via RSS